TRANSACTIONS OF THE SECTIONS. 45 



the lowest atomic weight of lead is 2-62 ; yet this difference, great as 

 it seems, results from a difference of only one 74th of a grain in the 

 lead left from 100* grains of the protoxide. Considering that this dif- 

 ference represents the widest variation that occurred in the experi- 

 ments, I need not say how much accuracy the smallness of its extent 

 implies. It is calculation that so much magnifies this inconsiderable 

 experimental difference, and in two ways — first, whatever is subtracted 

 from the lead comes to be added to the oxygen, and the converse; and 

 second, the difference in the oxygen comes, when computed on the 

 lead, to be multiplied by about thirteen times, the atomic weight of 

 lead being thus much greater than the atomic weight of oxygen ; so 

 that, in these two ways, an error of only one in 7400* on the protoxide 

 increases to 2-62 on 1394* of the protoxide, or on 1294* of lead, or to 

 about one in every 500" of the metal. Thus the atomic weight of lead 

 comes to vary in a ratio fifteen times as much as the experiment 

 whence it is derived. Certainly it is an unhappy circumstance when 

 calculation so much magnifies the unavoidable errors of experiment, 

 more especially when, as will be seen presently to happen in the 

 instance of lead, the original variation comes to be complexified with 

 other variations, and the error in atomic weight comes to be commu- 

 nicated in aggravated amount from element to element. 



The Sulphate and the Nitrate of Lead. 



Berzelius converted known weights of pure lead into a solution of 

 the nitrate, and, adding thereto an excess of dilute sulphuric acid, he 

 evaporated the whole to dryness, and heated the mass to redness, so as 

 to obtain the dry sulphate of lead. Turner repeated the experiment 

 nearly in the same manner ; but he made allowances for some slight 

 corrections that were neglected by Berzelius. The weight acquired by 

 the lead, in becoming the sulphate, should be increased on account of 

 the buoyancy of the air, and should be diminished on account of the 

 solid impurities of the acids employed. 



On 100' of lead the correction for buoyancy is -f-0125 



for acid impurities —•0190 (Turner.) 



Nett correction .... —-0065 



Making allowance for this correction in the experiments of Berzelius, 

 100- of lead afforded of the sulphate, 



Berzelius. Tui-ner. 



146-3741 Tv/r 146-4301 ,;. 



146-396 \ , f^l"^, 146-398 \ , ^^f"^, 



146-433 J 1*^*01 146-375] l^^'^^l 

 146-451 



The experiments of Turner, taken along with the three first of Ber- 

 zelius's, would seem to indicate that the last experiment of Berzelius's 

 affords too high an increase. From a comparison of the results ob- 

 tained by both chemists, the following quantities have been adopted in 



